Skip to content

Actions: facebookresearch/fairscale

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
111 workflow runs
111 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

add context parallel group init to mp init pre-commit #720: Pull request #1174 synchronize by amylittleyang
April 3, 2024 01:05 34s add_context_parallel_init
add_context_parallel_init
April 3, 2024 01:05 34s
add context parallel group init to mp init pre-commit #719: Pull request #1174 opened by amylittleyang
April 3, 2024 01:03 38s add_context_parallel_init
add_context_parallel_init
April 3, 2024 01:03 38s
Avoid calling _free_fp16_param_shard() too early pre-commit #718: Pull request #1164 synchronize by awgu
April 1, 2024 18:09 17s ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
April 1, 2024 18:09 17s
Added reshard hook for frozen params in backward pre-commit #716: Pull request #1159 synchronize by awgu
March 27, 2024 23:33 14s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
March 27, 2024 23:33 14s
Changed to only run reshard hook if all gradients computed pre-commit #715: Pull request #1166 synchronize by awgu
March 27, 2024 23:33 16s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
March 27, 2024 23:33 16s
Added reshard hook for frozen params in backward pre-commit #714: Pull request #1159 synchronize by awgu
March 27, 2024 23:29 14s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
March 27, 2024 23:29 14s
Changed to only run reshard hook if all gradients computed pre-commit #713: Pull request #1166 synchronize by awgu
March 27, 2024 23:29 17s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
March 27, 2024 23:29 17s
Avoid calling _free_fp16_param_shard() too early pre-commit #712: Pull request #1164 synchronize by whbldhwj
March 25, 2024 18:56 18s ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
March 25, 2024 18:56 18s
Added requires_grad check for params_with_grad method pre-commit #711: Pull request #1171 opened by whbldhwj
March 25, 2024 18:54 18s ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard_jie_fix
ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard_jie_fix
March 25, 2024 18:54 18s
Fairscale support for only performing allreduce in last microbatch pre-commit #710: Pull request #1168 synchronize by jiecaoyu
March 22, 2024 09:50 17s ngoyal_changes_for_pp_fp8_jiecaoyu_norm_allreduce
ngoyal_changes_for_pp_fp8_jiecaoyu_norm_allreduce
March 22, 2024 09:50 17s
Bump black from 22.3.0 to 24.3.0 pre-commit #709: Pull request #1169 opened by dependabot bot
March 20, 2024 16:58 37s dependabot/pip/black-24.3.0
dependabot/pip/black-24.3.0
March 20, 2024 16:58 37s
Fix params_with_grad in FSDP when the model has frozen parameters pre-commit #708: Pull request #1167 opened by whbldhwj
March 4, 2024 18:25 20s ngoyal_changes_for_pp_fp8_akadian_jie_fix
ngoyal_changes_for_pp_fp8_akadian_jie_fix
March 4, 2024 18:25 20s
Changed to only run reshard hook if all gradients computed pre-commit #707: Pull request #1166 opened by awgu
February 29, 2024 15:58 1m 5s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
February 29, 2024 15:58 1m 5s
Added reshard hook for frozen params in backward pre-commit #706: Pull request #1159 synchronize by awgu
February 29, 2024 15:55 18s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
February 29, 2024 15:55 18s
Avoid calling _free_fp16_param_shard() too early pre-commit #705: Pull request #1164 synchronize by jiecaoyu
February 21, 2024 11:39 17s ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
February 21, 2024 11:39 17s
Avoid calling _free_fp16_param_shard() too early pre-commit #704: Pull request #1164 opened by jiecaoyu
February 20, 2024 02:17 18s ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
ngoyal_changes_for_pp_fp8_jiecaoyu_free_fp16_shard
February 20, 2024 02:17 18s
Assign self.norm_type to input norm_type pre-commit #703: Commit 9a173bf pushed by brad-mengchi
February 2, 2024 23:52 36s main
main
February 2, 2024 23:52 36s
Should assign norm_type instead of scale_grad_by_freq pre-commit #702: Pull request #1162 opened by brad-mengchi
February 2, 2024 23:42 37s brad-mengchi:main
brad-mengchi:main
February 2, 2024 23:42 37s
added option for no PG validation for faster init pre-commit #701: Pull request #1161 opened by ngoyal2707
January 22, 2024 15:14 22s ngoyal_changes_for_pp_fp8_no_pg_validation
ngoyal_changes_for_pp_fp8_no_pg_validation
January 22, 2024 15:14 22s
Added reshard hook for frozen params in backward pre-commit #699: Pull request #1159 synchronize by awgu
January 12, 2024 19:27 19s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
January 12, 2024 19:27 19s
Added reshard hook for frozen params in backward pre-commit #698: Pull request #1159 opened by awgu
January 12, 2024 19:23 16s awgu:ngoyal_changes_for_pp_fp8_awgu
awgu:ngoyal_changes_for_pp_fp8_awgu
January 12, 2024 19:23 16s
[not to be merged yet] added temp changes for fp32 main grad, might not work for TE pre-commit #695: Pull request #1151 synchronize by ngoyal2707
December 7, 2023 01:27 20s ngoyal_temp_fp32_main_grad
ngoyal_temp_fp32_main_grad
December 7, 2023 01:27 20s
fixed broken clipping pre-commit #694: Pull request #1154 opened by ngoyal2707
December 7, 2023 01:27 16s ngoyal_main_grad_fix
ngoyal_main_grad_fix
December 7, 2023 01:27 16s
[not to be merged yet] added temp changes for fp32 main grad, might not work for TE pre-commit #693: Pull request #1151 synchronize by jiecaoyu
December 6, 2023 23:51 19s ngoyal_temp_fp32_main_grad
ngoyal_temp_fp32_main_grad
December 6, 2023 23:51 19s
fix .grad=None issue when param is not sharded pre-commit #692: Pull request #1153 opened by jiecaoyu
December 6, 2023 01:16 19s jiecaoyu_temp_fp32_main_grad_fix
jiecaoyu_temp_fp32_main_grad_fix
December 6, 2023 01:16 19s